|Publication number||US20060093019 A1|
|Application number||US 11/250,371|
|Publication date||May 4, 2006|
|Filing date||Oct 17, 2005|
|Priority date||Oct 15, 2004|
|Also published as||EP1648093A1, EP1648093B1, US7924904|
|Publication number||11250371, 250371, US 2006/0093019 A1, US 2006/093019 A1, US 20060093019 A1, US 20060093019A1, US 2006093019 A1, US 2006093019A1, US-A1-20060093019, US-A1-2006093019, US2006/0093019A1, US2006/093019A1, US20060093019 A1, US20060093019A1, US2006093019 A1, US2006093019A1|
|Inventors||Rohit Gaikwad, Jason Trachewsky|
|Original Assignee||Broadcom Corporation|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (9), Referenced by (6), Classifications (7), Legal Events (4)|
|External Links: USPTO, USPTO Assignment, Espacenet|
This application claims priority of U.S. Provisional Patent Application Ser. No. 60/618,791, filed on Oct. 15, 2004. The subject matter of the earlier filed application is hereby incorporated by reference.
1. Field of the Invention
The present invention relates to a system and method for cancelling harmonics of a digital system clock frequencies when the harmonics are coupled with an incoming signal.
2. Description of the Related Art
In any digital integrated circuit or system there is typically a running digital clock. Non-linear effects in the system typically cause the base digital clock frequency to lead to harmonics of that clock frequency. For example, if the base clock sequence is 100 MHz, some processes on the chip may cause undesirable components of 10 to 20 harmonics that can occur at 1 MHz and 2 MHz respectively and 1 GHz and 2 GHz respectively from the 100 MHz clock. These undesirable harmonics are also known as spurs. In a particular system, one can accurately predict the exact frequencies of the spurs because of the known base clock and harmonic frequency.
In a fast band system, i.e., a system that is operating at a high frequency with a fast band signal around the center carrier frequency, the fast band signal is down converted to base band by a radio chip. A base band processor then processes the base band signal. Specifically, the base band signal is digitized and a digital signal processor processes the base band signal. The digital signal processor usually works the running digital clocks. Therefore, when the RF section and the digital section are put on the same chip, the incoming digitized signal and the clock signal can couple easily. As such, the digital clock may create spurs all the way up to the fast band systems region of operation. Therefore, at the input of an RF, a desired signal or packet will be coupled with the spur frequency. The desired signal is typically a spread out signal, for example, 20 MHz in bandwidth. The spur frequency, on the other hand, is a very narrow band which is determined by the harmonic of the clock frequency.
A typical method of suppressing harmonics or spurs is through the use of electrical isolation between the digital system clock and the RF circuitry. However, isolation is not completely successful, and the in-band spurs can create numerous problems including carrier detection problems such as false detection. Additionally, the spurs created by the harmonics can degrade the signal-to-noise ratio. As the signal-to-noise ratio decreases, there is usually a higher incidence of dropped packets.
According to one aspect of the invention, there is provided a network device for cancelling spurs without affecting an incoming signal. The network device includes an estimator for estimating amplitude and phase of a spur over a predetermined period of time. The network device also includes processing means for freezing further estimation of the amplitude and phase of the spur, for cancelling for an estimated spur and for allowing incoming packets. The network device further includes subtracting means for subtracting the estimated spur from an incoming packet, wherein the estimated spur is subtracted from the incoming packet without affecting incoming signals that are not part of the estimated spur.
According to another aspect of the invention, there is provided a method for cancelling spurs without affecting an incoming signal. The method includes the steps of estimating amplitude and phase of a spur over a predetermined period of time and freezing further estimation of the amplitude and phase of the spur. The method also includes the steps of cancelling for an estimated spur and allowing incoming packets. The method further includes the step of subtracting the estimated spur from an incoming packet, wherein the estimated spur is subtracted from the incoming packet without affecting incoming signals that are not part of the estimated spur.
According to another aspect of the invention, there is provided an apparatus for cancelling spurs without affecting an incoming signal. The apparatus includes estimating means for estimating amplitude and phase of a spur over a predetermined period of time and freezing means for freezing further estimation of the amplitude and phase of the spur. The apparatus also includes cancelling means for cancelling for an estimated spur and allowing means for allowing incoming packets. The apparatus further includes subtracting means for subtracting the estimated spur from an incoming packet, wherein the estimated spur is subtracted from the incoming packet without affecting incoming signals that are not part of the estimated spur.
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention that together with the description serve to explain the principles of the invention, wherein:
Reference will now be made to the preferred embodiments of the present invention, examples of which are illustrated in the accompanying drawings.
This invention therefore seeks to cancel harmonics through the use of estimation circuitry which is configured to remove the spur without affecting the signal. According to an embodiment of the inventive system, since the clock frequency is known, it is relatively easy to determine exact frequencies of the spurs relative to the channel carrier frequencies. For example, if the clock frequency is 80 MHz and if the center RF frequency is 2.412 GHz, it is possible to determine that the in-band harmonics will occur at 2.416 GHz. In a wireless communication scenario such as, for example, 802.11g, there are 14 different channels or “slices” of the 2.4 GHz band. Using the example of an 80 MHz digital clock frequency, the harmonics for each of the 14 channels can easily be determined, and appropriate estimation and filtering may be applied to remove the spurs.
According to an embodiment of the invention, estimator 104 performs an estimation of the spur's amplitude and phase over a short period of time in the absence of incoming signals/packets. The estimation of the spur amplitude and phase may be repeated over a period of time. For every time step, the estimation, Ak, is obtained according to the following formula:
A k+1 =A k −μ·∂[e k H ·e k ]/∂A k
A k+1=(1−μ)·A k +μ·e i.2.π.k.f0/fs ·x k
wherein Ak is the estimation of the spur amplitude and phase;
xk is the incoming signal;
ei.2.π.k.f0/fs is an exponential of the harmonic at the frequency, f0;
fs is the sampling frequency;
μ, a register variable for adjusting the tracking rate of the spurl.
After obtaining an initial estimation, processor 106 multiplies the initial estimation by ei.2.π.k.f0/fs, which is an exponential of the harmonic at the frequency, f0. At some time thereafter, processor 106 subtracts the product from xk, an incoming signal.
Specifically, estimator 104 tracks the signal within a narrow band of energy, f0. As such, estimator 104 is a narrow adaptive filter that only lets in energy at the spur harmonic, f0, and blocks out all other energy. Thus, when the estimation is being performed, no signal energy is allowed to enter estimator 104, only spur energy is allowed to enter estimator 104. This provides a good estimation of the spur, even in the presence of noise since only the noise at the spur frequency is allowed to enter the filter.
As mentioned above, estimation of the spurs can occur, for example, at a time when there are no packets being transmitted or received, i.e., in the absence of signal. As is known to one skilled in the art, the spurs will exist in the absence of a signal since the digital clock is running on the substrate. As is also known to those skilled in the art, every packet reception is followed by a period of silence which allows estimator 104 to estimate the amplitude and phase of the spur in the absence of a signal. After estimating the spur amplitude and phase, spur canceller 100 freezes the estimation. Thereafter, spur canceller 100 starts cancelling for the spur. Once a packet is received, spur canceller 100 subtracts the estimated spur from the packet. Since the packet is expected to have the effect of the spur as it comes into ADCs 102, a correct estimation and subtraction would result in spur cancellation without affecting the incoming packet information at the same frequency as the spur.
While the spur amplitude and phase may be estimated in the absence of incoming signal, it is difficult to guarantee the absence of a signal during the estimation. As is known to those skilled in the art, one may rely on higher network directives, such as gaps between packets to determine the absence of signal; however, this method may be error prone. Therefore, in embodiments where there are no means of estimating the spur amplitude and phase in the absence of a signal, estimator 104 may estimate the amplitude and phase of the spur all the time at the same adaptive rate. As such, there is no initial estimation of the spur amplitude and phase. Therefore, during cancellation, spur canceller 100 will always notch out energy at the spur frequency. Even though in this embodiment the portion of the signal at the same frequency as the spur will also be notched out, it should be noted that the signal loss is minimal.
The above discussion assumes that there is no change in the gain in an RF front end where the signal is down converted from fast band to baseband and gained up appropriately. Currently, if there is a small incoming signal, the RF front end gains it up so that ADCs 102 can see the full range of the signal and if the signal is too large, it attenuates the signal properly so that it fits within the ADC range. In situations where there is a change in gain as packets are received on the front end, the estimation must be updated to reflect the gain changes. According to the embodiment outlined above, before the packet enters spur canceller 100 and the gain is determined, estimator 104 has already estimated the amplitude and phase of the spur and the spur canceller 100 is already cancelling for the spur. When the packet enters spur canceller 100, from the preamble of the packet, spur canceller can determine the appropriate gain for the packet and automatically apply the gain to the packet. The gain changes are determined by the packet received strength. It should be noted that if the desired signal strength changes, the gain for the spur will also change. As such, estimator 104 has to account for the gain change during packet reception. So if an estimate is based upon a level of, for example, 70 dB, but the packet strength coming in is 30 dB, then the spur value will be lower and the estimation needs to be updated to reflect this change in signal strength. As such, in this embodiment of the invention, spur canceller 100 adjusts the estimated component by the gain delta from when the intial estimation occurred to the packet gain that is being applied to the packet.
During gain change, the phase response may also change. Thus, to determine the phase change, spur canceller 100 runs amplitude and phase estimation during the packet reception with a slow μ, a variable for tracking the spur. In an embodiment of the invention, spur canceller 100 starts tracking the spur over the packet slowly. From the preamble to the payload portion of the packet, spur canceller 100 can track out what the phase of the packet was and determine the phase change.
One can imagine that after the initial estimation of the amplitude and phase of the spur, there can be a gain change in the front end. To account for this, spur canceller 100 may continue to run estimator 104 at a slow adaptive rate of μ. It should be noted that the adaptive rate of μ may be zero if necessary. So that as the gain and phase change, the estimator 104 will update and hopefully only track the gain and phase from the spur and not the signal energy.
Embodiments of the above mentioned invention may be implemented in an RF-Baseband integrated chip.
In Step 2030, estimator 104 performs an estimation of the spur's amplitude and phase over a short period of time and in the absence of signal. In Step 2040, spur canceller 100 then freezes the estimation and starts cancelling for the spur before allowing packets to enter. In Step 2050, once a packet is received, estimator 104 subtracts the estimated spur from the packet. In Step 2060, where there is no means of estimating the amplitude and phase of the spur in the absence of a signal, estimator 104 may estimate the amplitude and phase of the spur all the time at an adaptive rate. It should be noted that the tracking rate μ can possible be varied over the length of the packet. For example, there may be different tracking rates during silence period, preamble and payload/data portion of the packet. In Step 2070, spur canceller 100 may adjust the estimated component by the gain delta from when the estimation occurred to the packet gain that is being applied to the packet. In Step 2080, to determine the phase change, spur canceller 100 runs amplitude and phase estimation during the packet reception with a slow μ for tracking the spur. The steps above may be repeated after every packet or after a predetermined period of time. Since the spur strength can be a function of temperature on the substrate, if the temperature changes then the estimate of the spur amplitude and phase may be stale. Therefore, the predetermined period of time can be set to account for activities that may affect the estimate of the spur amplitude and phase.
In one embodiment, therefore, the invention includes a spur canceller having an estimator for estimating spur magnitude and a phase, and a subtracter for subtracting an estimated spur value from an incoming packet at an appropriate time. The invention can also include an updater for updating the estimation based upon the signal level of an incoming packet, to dynamically adjust for varying spur magnitudes. The invention can be applied to any configuration where a fixed clock signal is creating a harmonic or a spur on an adjacent circuit, and more particularly to be applied to wireless RF communications where a digital clock pulse is operating in conjunction with RF frequencies.
The foregoing description has been directed to specific embodiments of this invention. It will be apparent, however, that other variations and modifications may be made to the described embodiments, with the attainment of some or all of their advantages. Therefore, it is the object of the appended claims to cover all such variations and modifications as come within the true spirit and scope of the invention.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US4207543 *||Jul 18, 1978||Jun 10, 1980||Izakson Ilya S||Adaptive filter network|
|US4885587 *||Dec 22, 1988||Dec 5, 1989||Westinghouse Electric Corp.||Multibit decorrelated spur digital radio frequency memory|
|US5687735 *||Mar 28, 1996||Nov 18, 1997||Hewlett-Packard Company||Robust time-diversity filter and method for removing electromagnetic interference|
|US6384589 *||Apr 4, 2001||May 7, 2002||Tektronix, Inc.||Reference frequency spur cancellation in synthesized measurement receivers|
|US6563931 *||Jul 29, 1992||May 13, 2003||K/S Himpp||Auditory prosthesis for adaptively filtering selected auditory component by user activation and method for doing same|
|US6570864 *||Nov 16, 1999||May 27, 2003||Electronics And Telecommunications Research Institute||Integrated receiving apparatus of subtractive interference cancellation receiver and adaptive MMSE receiver|
|US20020160734 *||Dec 21, 2001||Oct 31, 2002||Tao Li||Direct conversion receiver architecture|
|US20030040292 *||Dec 10, 2001||Feb 27, 2003||Peterzell Paul E.||Local oscillator leakage control in direct conversion processes|
|US20030176173 *||Mar 14, 2002||Sep 18, 2003||Nikolaus Klemmer||Direct automatic frequency control method and apparatus|
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US7646704 *||Oct 31, 2006||Jan 12, 2010||Motorola, Inc.||Method and apparatus for spur cancellation in an orthogonal frequency division multiplexing communication system|
|US8149896 *||Jan 4, 2006||Apr 3, 2012||Qualcomm, Incorporated||Spur suppression for a receiver in a wireless communication system|
|US8238480||Oct 5, 2007||Aug 7, 2012||Motorola Mobility Llc||Adaptive self-quieter suppression for OFDM wireless communication systems|
|US8817925 *||Mar 16, 2012||Aug 26, 2014||Qualcomm Incorporated||System and method for RF spur cancellation|
|US9112748 *||Feb 12, 2013||Aug 18, 2015||Qualcomm Incorporated||Reduction of small spurs in transmitters|
|US20130208826 *||Feb 12, 2013||Aug 15, 2013||Qualcomm Incorporated||Reduction of small spurs in transmitters|
|Cooperative Classification||H04B15/06, H04B2215/064, H04B15/02|
|European Classification||H04B15/02, H04B15/06|
|Jan 13, 2006||AS||Assignment|
Owner name: BROADCOM CORPORATION, CALIFORNIA
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:GAIKWAD, ROHIT;TRACHEWSKY, JASON ALEXANDER;SIGNING DATESFROM 20051220 TO 20060105;REEL/FRAME:017469/0401
|Nov 21, 2014||REMI||Maintenance fee reminder mailed|
|Apr 12, 2015||LAPS||Lapse for failure to pay maintenance fees|
|Jun 2, 2015||FP||Expired due to failure to pay maintenance fee|
Effective date: 20150412